MedTAG is an open-source biomedical annotation tool for diagnostic reports. MedTAG provides effective and intuitive tools to tag biomedical concepts contained in clinical reports.
The tool is available at this link (credentials are required).
A collaboration between users is enabled, and there are many valuable features to see report translations, annotations from other users, and user stats. Other features are:
- a web-based collaborative annotation platform with support for users and roles
- support for click-away mention annotation
- support for mentions highlighting in different colors
- automatic saving every time an action is performed
- sorting of medical reports according to two different strategies: lexicographic order and “unannotated-first” policy
- web responsive design to support mobile devices
- download of annotations and ground truths in several formats (i.e., BioC/XML, CSV, JSON)
- support for multi-label annotation
- support for document-level annotations
- multilingual support
- support for ontologies/concepts to use for the annotation process
- support for schema configuration, so that users can easily import data (i.e., reports, labels and concepts), as CSV files, and choose which report fields to annotate.
- support for automatic annotation of all the annotation types for reports belonging to three use cases: colon, uterine cervix and lung. Automatic annotation is available for english reports
- support for annotation of PubMed articles
- support for visualization of report’s annotations
- support for inter-annotator agreement
- support for the upload and visualization of other team members’ ground-truths
Everything is well documented at this link, where the tool is also available in a Docker version.
A paper describing MedTAG has been accepted on BMC Medical Informatics & Decision Making.
We also developed an alternative version of MedTAG called DocTAG for annotating documents in typical information retrieval evaluation campaigns. A video describing the tool is available here.
